Since substantial neuronal loss has already occurred by the time Alzheimer's Disease (AD) manifests clinically, considerable research has been directed towards developing biomarkers for the reliable diagnosis of mild AD and the prodromal stage of mild cognitive impairment. The best studied biomarkers, A1-42 peptide and tau/P-tau, can distinguish cohorts of moderate versus mild AD or MCI subjects, but values vary considerably on an individual basis. Therefore, there is an urgent need for biomarkers with greater specificity and sensitivity. Here we propose to use our proprietary yeast display antibody library to differentially screen for antibody clones that bind to antigens whose structure/epitopes are markedly altered in AD compared to age-matched control CSF. The yeast clones harvested by FACS will be used to generate a yeast cell-based microarray to probe individual AD and normal subject CSF samples to identify those clones that exhibit marked differences in binding to AD CSF samples. The most promising AD-specific antibody clones will be used to immuno-enrich their cognate antigen for LC-MS/MS identification. This feasibility study will provide the groundwork for an expanded antigen identification effort in Phase II, followed by a more extensive assessment of the diagnostic potential of putative AD-specific antibody-antigen pairs.